search

PolicyEngine / policyengine-us

The PolicyEngine US Python package contains a rules engine of the US tax-benefit system, and microdata generation for microsimulation analysis.
https://policyengine.org/us
GNU Affero General Public License v3.0
100 stars 174 forks source link

Baseline poverty rate is too low (6.8%) #2036

Closed MaxGhenis closed 1 year ago

MaxGhenis commented 1 year ago

This must have just changed, as the ARPA EITC post shows 9.6% baseline SPM. Could it have been the survey-enhance transition?

This shows up in the app, but locally (from master) I get:

In [1]: from policyengine_us import Microsimulation

In [2]: b = Microsimulation()

In [3]: b.calc("in_poverty").mean()
Out[3]: 0.12453627661737537
MaxGhenis commented 1 year ago

I added a new CPS 2023 data release based on my locally generated file, which correctly produces a 9.6% poverty rate locally. But after updating the API, it's still 6.8% baseline in the app. @nikhilwoodruff are any other steps needed to update the microdata?

nikhilwoodruff commented 1 year ago

Could you verify if setting state tax to reported produces the lower poverty rate locally?

MaxGhenis commented 1 year ago

Oh I'm seeing now that we need to point to an asset ID:

https://github.com/PolicyEngine/policyengine-us/blob/a796026916c14a433bbc1269f8348fd3f118495e/policyengine_us/data/datasets/cps/cps.py#L345-L352

How do I make that for a new file?